What causes Database Error: Redshift Connection Timeout?

Redshift cluster idle session timeout (default 1 hour) closing dbt's connection mid-run. WLM queue timeout killing long-running dbt queries before completion. VPC security group or network ACL blocking the dbt runner's IP address. Redshift cluster in maintenance mode or resizing. TCP keepalive not configured on the dbt host, causing idle connections to be dropped by NAT/firewall

How do I fix Database Error: Redshift Connection Timeout?

Set `keepalives_idle: 60` in the Redshift profile in profiles.yml to prevent idle connection drops. Increase the Redshift WLM query timeout for the queue used by dbt. Configure the Redshift parameter group: increase `statement_timeout` for long dbt models. Verify network connectivity: run `psql -h -U -d -p 5439 -c 'SELECT 1'` from the dbt runner. If using a VPN or Bastion, ensure TCP keepalive is also configured on the network appliance. Check the Redshift system table `stl_connection_log` for connection termination events

High severityconnectivity

Power BI Refresh Error:
Database Error: Redshift Connection Timeout

What does this error mean?

dbt failed to connect to Amazon Redshift, or a running query was terminated due to a connection timeout. This can be a network timeout, an idle session timeout, or a WLM (Workload Management) queue timeout.

Common causes

1Redshift cluster idle session timeout (default 1 hour) closing dbt's connection mid-run
2WLM queue timeout killing long-running dbt queries before completion
3VPC security group or network ACL blocking the dbt runner's IP address
4Redshift cluster in maintenance mode or resizing
5TCP keepalive not configured on the dbt host, causing idle connections to be dropped by NAT/firewall

How to fix it

1Set `keepalives_idle: 60` in the Redshift profile in profiles.yml to prevent idle connection drops
2Increase the Redshift WLM query timeout for the queue used by dbt
3Configure the Redshift parameter group: increase `statement_timeout` for long dbt models
4Verify network connectivity: run `psql -h <cluster_endpoint> -U <user> -d <db> -p 5439 -c 'SELECT 1'` from the dbt runner
5If using a VPN or Bastion, ensure TCP keepalive is also configured on the network appliance
6Check the Redshift system table `stl_connection_log` for connection termination events

Frequently asked questions

Where is `keepalives_idle` configured for the Redshift dbt adapter?

In your `profiles.yml` under the Redshift target: add `keepalives_idle: 60` (seconds). This configures the TCP keepalive to send probes every 60 seconds on idle connections.